1. Field of the Invention
The present invention relates generally to a semiconductor element, a semiconductor device having the semiconductor element, and methods for producing the semiconductor element and the semiconductor device. More specifically, the invention relates to a semiconductor element of a high-frequency discrete transistor, a semiconductor device, to which the semiconductor element is bonded, and methods for producing or fabricating the semiconductor element and the semiconductor device.
2. Description of the Prior Art
FIG. 13 shows a cross section of a typical discrete bipolar transistor. As shown in FIG. 13, the discrete bipolar transistor comprises base electrodes B, B and an emitter electrode E, which are formed on the surface of a semiconductor device, and a collector electrode C formed on the reverse surface of the semiconductor device. If such an npn transistor is used to amplify a high-frequency signal, there is a problem in that a collector-to-base parasitic capacity C increases, so that it is not possible to obtain sufficient high-frequency characteristics. In order to eliminate this problem, in recent years, a bipolar transistor shown in FIG. 11 has been used. In this bipolar transistor, a collector electrode C is also formed on the surface of a semiconductor device. Thus, if the collector electrode C is formed on the surface of the semiconductor device, there is an advantage in that it is possible to decrease the collector-to-base parasitic capacity C.
FIG. 11 is a schematic sectional view of a semiconductor device, to which a semiconductor element of a high-frequency npn transistor is bonded, and FIG. 12 is a plan view of the semiconductor device.
As can be seen particularly from FIG. 11, an emitter electrode E, a base electrode B and a collector electrode C are provided on the top of the npn transistor. These electrodes E, B, C are connected to an emitter lead ER, a base lead BR and a collector lead CR via an emitter bonding wire EW, a base bonding wire BW and a collector bonding wire CW, respectively. In FIG. 11, the bonding wires BW, CW and bonding pads BP, CP, which will be described later, are omitted.
More specifically, as can be seen from FIG. 12, the emitter electrode E is connected to emitter bonding pads EP, EP, the base electrode B is connected to a base bonding pad BP, and the collector electrode C is connected to a collector bonding pad CP. The emitter bonding pads EP are connected to the emitter lead ER by means of the emitter bonding wires EW. The base bonding pad BP is connected to the base lead BR by means of the base bonding wire BW. The collector bonding pad CP is connected to the collector lead CR by means of the collector bonding wire CW. One end of the base lead BR, one end of the collector lead CR, and both ends of the emitter lead ER are exposed to the outside of a package PG so as to be capable of being connected to a wiring substrate or the like.
As described above, in the conventional semiconductor device, the electrodes E, B, C of the npn transistor are connected to the lead ER, BR, CR by means of the bonding wires EW, BW, CW, respectively. However, these bonding wires EW, BW, CW have excessive impedance, which causes the deterioration of the high-frequency characteristics of the npn transistor. That is, an inductance is produced under the influence of magnetic fields caused by the bonding wires EW, BW, CW themselves, and this inductance serves as excessive impedance, which causes the deterioration of the high-frequency characteristics of the npn transistor, i.e., the deterioration of the current amplification in a high frequency band. In particular, if the npn transistor is used for an emitter grounded circuit, the influence of impedance of the emitter bonding wire EW out of the bonding wires EW, BW, CW is greatest. Also as can be seen from FIG. 12, the emitter bonding pad EP and the emitter lead ER are typically connected to each other by means of two wires. Therefore, the influence of the magnetic field cause by the emitter bonding wires EW is far greater than that of the bonding wires BW, CW.
Therefore, the emitter electrode E may be provided on the reverse surface of the semiconductor device, and the base electrode B and the collector electrode C may be provided on the surface of the semiconductor device. However, if the emitter electrode E is provided on the reverse surface of the semiconductor device, the emitter electrode E is formed on the whole reverse surface of the semiconductor device, so that the area of the emitter electrode E increases. If the area of the emitter electrode E increases, the switching takes a lot of time, so that it is not possible to enhance the high-frequency characteristics. That is, if the area between the emitter and base is large, there is a problem in that the charge and discharge for carriers take a lot of time, so that it is not possible to increase the switching rate. Therefore, in order to enhance the high-frequency characteristics, it is not desired to provide the emitter electrode E directly on the reverse surface of the semiconductor device.